Anal. Chem. 1986, 58, 2414-2420
2414
Table 111. Sample Data % by volume
sample
EPA
EPA-EOD-1 EPA-EOD-2 EPA-EOD-3 EPA-EOD-4 EPA-EOD-5 EPA-EOD-6 EPA-EOD-7 EPA-EOD-8 EPA-EOD-9 EPA-EOD-10
16.01 0.35" 9.87 f 0.11 19.46 f 0.46 NDb 4.55 f 0.30 0.44 0.04 8.37 f 0.46 1.18 f 0.03 10.72 0.32 16.15 f 0.52
* *
*
ITDS 16.10 f 0.04" 9.76 f 0.26 19.55 f 0.42 ND 4.88 f 0.08 0.33 f 0.04 8.32 f 0.13 1.26 f 0.15 10.84 f 0.27 16.17 f 0.17
% diff
0.56 1.11 0.46 7.26 25.00 0.60 6.78 1.12 0.13
"95% confidence limit. *ND, not detected.
molecular fraction of CH3CHOH+. By use of both the 31 and 45 m l e ratios the overall sensitivity and the detection limits were improved. Figure 2b,c is only shown for explanatory purposes, Figure 2a is the chromatogram used for quantification of the alcohol content of the gasoline. As indicated by Figure 2b,c, ethanol is present in both MID chromatograms. By use of both mle ratios, 31 and 45, ethanol could be determined a t the upper limit of 20% total volume. Twenty percent methanol or ethanol content was used as the upper limit since this value is well above the legal limit set by the Ohio Attorney General's Office. The remaining six alcohols are predominately found in concentrations of less than 1% since they are most commonly used as cosolvents for the other alcohols. T o evaluate the method, gasoline samples that had been assayed by the EPA method were reanalyzed by the GC-ITDS
procedure. Included in these samples were ten samples containing ethanol in concentrations up to 20% by volume that had been prepared for our laboratory by the EPA. The values that were obtained for the ten EPA samples by both methods are contained in Table 111.
ACKNOWLEDGMENT We wish to thank James Petroff of the Consumer Fraud Division of the Ohio Attorny General's Office for providing the gasoline samples used in the testing of this method. Registry No. Methanol, 67-56-1;ethanol, 64-17-5;2-propanol, 67-63-0; 1-propanol, 71-23-8; 2-methyl-2-propano1, 75-65-0; 2butanol, 78-92-2; 2-methyl-1-propanol,78-83-1; 1-butanol,71-36-3.
LITERATURE CITED Thomas, K. The Aviation Consumer 1983, 13, 15-21. Renzoni, G. E.; Shankland, E. G.; Gaines, J. A,; Callis, J. B. Anal. Chem. 1985, 57, 2864-2867. "Ohlo Consumer Sales Practices Act and Substantive Rules", Ohio Attorney General's Office, pp 48-49. Pauls, R. E.; McCoy, R. W. J. J . Chromatogr. Sci. 1981, 79, 558-561. Sevick, J. HRC CC, J . High Resolut. Chromatogr. Chromatogr. Commun. 1980, 3 , 166-168. Durand, J. P; Petroff, N. Rev. Inst. F r . Pet. 1982, 3 7 , 575-578. Chem. Abstr. 1982, 9 7 , 112187a. Luke, L. A.; Ray, J. E. Analyst (London) 1984, 709, 989-992. Lockwood, A. F.; Craddock, B. D. Chromatographia 1983, 17, 65-68. Johansen, N. G. HRC CC , J High Resolut. Chromatogr Chromatogr . Commun. 1984, 8 , 487-489. Zinbo, M. Anal. Chem. 1984, 5 6 , 244-247. US EPA NEIC 1984, Ohio Attorney General's Office. Silverstein, R. M., Bassler, G. C., Morrill, T. C. Spectrometric Identification of Organic Compounds, 4th ed.; 1981. Ion Trap Detector Operation Manual, Revision 6 ,Finnigan-MAT Corp., San Jose, CA.
.
RECEIVED for review March 18,1986.
.
Accepted June 1,1986.
Determination of Picogram per Cubic Meter Concentrations of Tetra- and Pentachlorinated Dibenzofurans and Dibenzo-p-dioxins in Indoor Air by High-Resolution Gas Chromatography/High-Resolution Mass Spectrometry R. M. Smith,* P. W. O'Keefe, D. R. Hilker, and K. M. Aldous Wadsworth Center for Laboratories and Research, New York State Department of Health, Empire State Plaza, Albany, New York 12201
An analytlcal method Is presented for the collectlon and quantlflcatlon of tetra- and pentachlorlnated dlbenrofurans (CDFs) and chlorinated dlbenro-pdbxlns (CDDs) In ambient alr. Thls method Is the first to deal with gaseous CDFs and CDDs as well as particulate-bound compounds and the first to validate and report the quantlflcatlon of these compounds at 1 pg/m3. Samples were collected with a glass flber filter followed by slllca gel contalned In a removable extraction thimble In a housing made of Teflon. The samples were Soxhiet extracted wlth benzene, cleaned up with alumlna and carbon adsorption llquld chromatography, and quantlfled with SP2330 caplllary GC/mass profile high-resolutlon mass spectrometry. A complex Isomer pattern of tetra- and penta-CDFs and tetra-CDDs lncludlng the 2,3,7,8-substltuted tetras was found in air samples taken from a contamlnated office building. Concentrations ranged from 0.23 pg/m3 2,3,7,8-tetra-CDD to 185 pg/m3 for total tetra-CDFs. Chlorlnated blphenylenes were also found. 0003-2700/86/0358-2414$01.50/0
Most of the interest concerning CDFs and CDDs in air has been related to incinerator stack emissions. Because no analytical method presently exists for the determination of picogram per cubic meter quantities of these compounds in indoor or outdoor air, risk assessment groups have been forced to use computer modeling or surrogate analyses (e.g., PCB) to estimate CDF and CDD concentrations. Although attempts have been made to analyze air particulate matter for CDFs and CDDs, no analytical method has dealt with the possibility of gaseous or desorbed compounds. Stack emission samples are commonly collected with a glass fiber filter followed by an XAD resin trap to collect both particulate and vapor-phase compounds. The filter and trap are contained in a complex water-cooled sampling train that is necessary to handle conditions of high temperature and moisture (1, 2); quantification is by GC/MS. For ambient air a simpler collection system can be used, but larger volumes (80 m3 or more) must be sampled. A high-volume ambient 0 1986 American Chemical Society
ANALYTICAL CHEMISTRY, VOL. 58, NO. 12, OCTOBER 1986
air sampler that can be applied to a variety of organics has been described ( 3 , 4 )that uses a glass fiber filter with polyurethane foam or other collection media. Because of the high toxicity (and low volatility) of certain 2,3,7,&substituted CDFs and CDDs, a prerequisite of any ambient air analytical method for these compounds is a detection limit of approximately 1 pg/m3 per isomer or better, based on criteria such as the New York State Health Department acceptable daily intake of 2,3,7,8-TCDD for humans of 2 (pg/kg)/day (5). Until now this had not been achieved. Attempts have been made to analyze air for CDFs/CDDs. For example, air samples collected on a filter after an industrial fire were analyzed by HRGC/HRMS; the results for TCDDs were inconclusive a t 5 pg/filter (6). Similarly no TCDDs or TCDFs could be detected by LRMS in several air samples taken following soot-producing incidents in Europe (7). T o obtain adequate detection limits for CDFs and CDDs in ambient air we developed a trapping system consisting of a glass fiber filter followed by a removable silica gel cartridge. Organic residues are readily removed from silica gel thereby minimizing background interferences such as encountered in trapping from air using polymer resins. Following extraction and cleanup of the sample, HRGC/HRMS provided selectivity to further eliminate background chemical noise and sample interferences which are commonly the limiting factors in obtaining low detection limits in analyses of this type (8). Our HRMS system featured a m a profile mode (HRMPM) which displays interfering compounds within a 100 mmu region but presently requires multiple injections for higher chlorinated congeners. We applied our method to the analysis of air samples taken after an electrical fire in 1981 in an office building in Binghamton, NY. The fire caused a 1060-gallon pyranol-filled transformer to overheat and spread chemical contamination (as black, oily soot) throughout the building (9, IO). Pyranol is a mixture of 65% Aroclor 1254 and 35% chlorinated benzenes which have been shown to pyrolyze producing polychlorinated dibenzo-p-dioxins (CDDs) and/or polychlorinated dihenzofurans (CDFs) (11, 12). Investigation of this and similar soot-producing incidents has resulted in the identification of CDFs, CDDs, chlorinated biphenylenes, and chlorinated pyrenes in these swts (13). GC/HRMS analysis of soot samples from the Binghamton building showed an average of 162,197,110,36, and 14 ppm of tetra- to octa-CDFs, respectively (14). 'One sample approximately 2.9 ppm 2,3,7,8-TCDD and 195 ppm 2,3,7,8-TCDF (1.5). After the building was extensively cleaned, it presented a unique opportunity for the analysis of residual picogram quantities of these compounds in the air.
EXPERIMENTAL SECTION Caution. Persons using the following procedures should be familiar with standard precautions for handling toxic CDDs and CDFs. Solvents. All solvents used were Burdick and Jackson "distilled in glass" except reagent grade acetone and toluene for preliminary glassware washing. Gases. Ultra-high-purity (99.999%) nitrogen (Liquid Carbon Corp., Albany, NY) was introduced for sample concentration with no purification trap via a Teflon line to a Pasteur pipet. Adsorbents. Silica Gel. The adsorbent for trapping CDFs and CDDs from air was silica gel 60 (EM Reagents, Darmstadt, FRG) 35-70 mesh. It was washed with 50% methanol in CH,CI,, filtered, oven dried a t 140 OC, and stored in a vacuum desiccator until packed into 25 X 85 mm Pyrex extraction thimbles (40-60 mesh frit, VWR Scientific, Rochester, NY). Aluminas. AG4 acid alumina and AG7 neutral alumina, both 100-200 mesh (Bio Rad, Richmond, CAI, were activated a t 450 OC in a muffle furnace overnight and stored in 8 X 250 mm (Ace Glass, Vineland, NJ) packed columns under vacuum for use within 24 h.
2415
Flgure 1. TCDD volatilization and trapping apparatus.
Carbon. Ten grams of PX-21 carbon adsorbent (Amoco Research, Naperville, IL) was mixed with 120 g of Celite 545, and the mixture was washed with benzene and CH,CI, and then dried in a vacuum oven (50 %/25 psig). Glass columns (8 X 85 mm, Ace Glass) were packed bottom to top with granular Na2S04,7 mm Celite 545,650 mg (25 mm) of carbon/Celite mixture, and again with Celite and Na,SO,. Tamping of layers with a glass rod was necessary. Glassware for Sample Concentration. One hundred milliliter round bottom flasks, 22 x 110 mm indented tapered borosilicate tubes (16),2 mm i d . x 10 cm Pyrex tubes, and l(t100 pL syringes (Hamilton 800 series, Reno, NV) were used. Cleaning. Between samples glassware was washed with acetone, toluene, and finally edistilled-in-glass" hexane. Triton X (2%) was used for washing Teflon samplers and as needed for glassware. Chromic acid solution was used as needed. Prior to use the Soxhlet apparatus was cycled with benzene for 1 h; the benzene was then discarded or checked for residual PCDD/PCDF. Syringes were disassembled for cleaning with the above solvents. Disposables. Boiling stones (Hengar, Philadelphia, PA), Pasteur disposable glass pipets, aluminum foil, hexane-washed glass wool, and high-purity 45 mm diameter EPM 2000 glass microfiber particulate filters ( c cn
z
w
c
z
n.fifi .. 0.81 0.72
ND ND ND total penta-CDDs eoncn Nn ma\ NE i0:ii
ND (0.5)
"Sample cleanup recovery for 2,3,6,7-tetra-CI biphenylene is 50%. As no standard is available, the recovery and MS response for pentachlorobiphenylenesare unknown. (clean ion source) and not obtainable on a daily basis; digital signal averaging was also applied. A more typical instrumental sensitivity was approximately 1-10 pglinjection. The contribution from laboratory contamination was assessed by analyzing duplicate method blanks: no detectable TCDDs were found. A method blank fortified to 1.0 pg/m3 was quantified at 0.71 pg/m3. Penta-CDDs quantified in samples 29-31 were not detected at 0.5 pg/m3. The results from a reinjection of several samples for tetra- and penta-
LBc in\ SCANS 161.174
100X~355B18
IODX~5961
"CONTROi"AI1
2.3.7.8-TCDF FORTIFIEDA11
u
50
0.88 0.90 1.38
total penta-CI hiphenylenesO concn 324/326
NUMBER O F C H L O R I N E S
total hexa-PCB m l e 360
303.8551
303.9013
319.8483
319.b964
303.9410
303.9013
3019470
z
s
339.5085
339.8596
339.9106
333.9335
3339836
MASS/Z Flg"rn 7. Mass profilesof (A) M+ of TCDF memod blank, (6) 2.3.7.ETCOF-fwtiRed ambient air, (C) control air, (D) (M blank. (E) TCDD method blank, and (F) 12 pg of "C TCDD standard.
+ 2)+ of penta-CDFs memcd
collection of the samples from the contaminated building. R e g i s t r y No, 2,3,7,8-TCDF, 51207-31-9; 2,3,7,8-TCDD, 1746-01-6; TCDD, 41903-57-5; TCDF, 30402-14-3; PCDF, 30402-15-4; hexa-PCB, 26601-64-9; tetrachlorobiphenylene, 26444-41-7.
LITERATURE CITED
I
'
, 1
50
117
,
, 5
~
i3,
,
)
23
I
L
I I o,
I I
) A:
1118mb
RETENTION TIME (minl/SCANS
, I33 9335
\. I31 3316
MASS/z
Flgwe 8. Capillary GC/HRMS data for tetraCWs: ion chromatograms for air sample 30 (A) M+ (B) (M -I- 2)'. (C) [U-'%]-2,3,7,8-TCDD internal standard, (D) mass profile for the same sample scans 83-95 showing the molecular ion of native 2,3,7,8-TCDD at m/z 319.8964, (E) the (M 2)' ion at m l z 321.8935 and mass-resolved pentachlorobiphenylene coeluter, and (F) [U-I3C]-2,3,7,8-TCDD internal standard.
+
chlorinated biphenylenes are shown in Table V. Hexachlorinated biphenyl (a major interferent in the analysis of tetrachlorobiphenylene after loss of two C1) was not observed at m l e 360. As most biphenylene standards were unavailable, quantitation was based on 2,3,6,7-TC1 biphenylene.
ACKNOWLEDGMENT The authors are grateful to Versar of Springfield, VA, for
Junk, G. J.; Jerome, B. A. Am. Lab. (Fairfield, Conn.) 1983, 75, 16-29. Stanley, J. S.;Haile, C. L.; Small, A. M.: Olson, E. P. EPA Contract No. 68-01-5915, January 1982. Lewis, R. G.; Brown, A. R.; Jackson, M, D. Anal. Chem. 1977, 49, 1668-1672. Lewis, R. G.; Jackson, M. D. Anal. Chem. 1982, 54, 592-594. Kim, N.; Hawley, J. Revised Risk Assessment-Binghamton State Office Building, June 7, 1963. HaNan, D. J.; Hass. J. R.; Schroeder, J. L.; Corbett, B. J. Anal. Chem. 1981, 53, 1755-1759. Rappe, C., personal communication. Lamparski, L. L.; Nestrick, T. J. Anal. Chem. 1980, 52, 2045-2054. Haughie, G. F.; Schecter. A. J.; Rothenberg, R. MMWR 1981, 30, 187-188, 193. Schecter, A. J. Chemosphere 1983, 12, 669-680. Buser, H. R.; Bosshard, H.: Rappe, C. Chemosphere 1978, 1, 109-1 19. Buser, H. R. Chemosphere 1979, 8, 415-424. Rappe, C.; Marklund, S.;Kjelier, L.; Bergquist, P.; Hansson, M. I n Chlorinated Dioxins and Dibenzofurans in the Total Environment II : Keith, L., Rappe, C., Choudhary, G., Eds.; Butterworth: Woburn, MA, 1985. O'Keefe, P. W.; Smith, R. M.; et al. EHP Environ. Health Perspect. 1985, 60, 201-209. Smith, R. M.; O'Keefe, P. W.; Hiiker, D. R.; Jelus-Tyror, 8. L.; Aldous, K. M. Chemosphere 1982, 8, 715-719. O'Keefe, P. W.; Meyer, C.; Dillon, K. Anal. Chem. 1982, 54, 2623-2625. O'Keefe, P. W.; Smith, R. M. I n ChlorinatedDioxins and Dibenzofurans in the Total Environment II; Keith, L., Rappe, C., Choudhury, G., Eds.; Butterworth: Worburn, MA, 1985. Mazer, T.; Hileman, F. D.; Noble, R. W.; Brooks, J. J. Anal. Chem. 1983, 55, 55-110. Olie, K.; Vermeulen, P. L.; Hutzinger. 0. Chemosphere 1977, 7, 455-459. Lustendhower, J. W. A.; Olie, K.; Hutzinger, 0. Chemosphere 1980, 9 , 501-522. Trace Chemistries of Fire; The Dow Chemical Co.: Midland, MI, 1978.
RECEIVED for review July 15, 1985. 1986. Accepted May 23, 1986.
Resubmitted March 17,